To my family

List of Papers

This thesis is based on the following papers, which are referred to in the text
by their Roman numerals.

Eriksson J., Larson G., Gunnarsson U., Bed'hom B., TixierBoichard M., Strmstedt L., Wright D., Jungerius A., Vereijken A.,
Randi E., Jensen P., Andersson L. (2008) Identification of the yellow
skin gene reveals a hybrid origin of the domestic chicken. PLoS
genetics 4(2): e1000010.

II

Rubin C.J., Zody M.C., Eriksson J., Meadows J.R., Sherwood E.,
Webster M.T., Jiang L., Ingman M., Sharpe T., Ka S., Hallbk F.,
Besnier F., Carlborg O., Bed'hom B., Tixier-Boichard M., Jensen P.,
Siegel P., Lindblad-Toh K., Andersson L. (2010) Whole-genome
resequencing reveals loci under selection during chicken
domestication. Nature 464: 587-591.

III Eriksson J., Hellstrm A.R., Rubin C.J., Wang C., Shumaila S.,
Kerje S., Gourichon D., Bedhom B., Tixier-Boichard M., Leif Andersson. (2010) A frameshift mutation in COMTD1 specifically
dilutes pheomelanin pigmentation in chicken. Manuscript.
Reprints were made with permission from the respective publishers.

Contents

Introduction ..................................................................................................... 9
Mutations the Source of Genetic Variation ........................................... 10
Functional Categorization of Mutations .............................................. 11
Identification of Sequence Elements Regulating Gene Expression ..... 13
Domestication ........................................................................................... 14
Domestication of the Chicken .............................................................. 16
Pigmentation ............................................................................................. 18
Carotenoid-Based Pigmentation .......................................................... 19
Melanin-Based Pigmentation ............................................................... 20
The Chicken Genome ............................................................................... 21
Methods for Genetic Mapping .................................................................. 22
Introduction .......................................................................................... 22
Linkage Analysis ................................................................................. 23
Identical-By-Descent Mapping ............................................................ 23
Mutation Detection .............................................................................. 25
Aims of this thesis ......................................................................................... 27
Present Investigations .................................................................................... 29
Mapping Monogenic Traits in Chicken .................................................... 29
Background .......................................................................................... 29
Results and Discussion ........................................................................ 31
Signatures of Selection in the Chicken Genome ...................................... 37
Background .......................................................................................... 37
Results and Discussion ........................................................................ 38
General Discussion and Future Perspectives ................................................ 43
Acknowledgements ....................................................................................... 47
References ..................................................................................................... 49

Abbreviations

ASIP
BCDO2
BCO2
bp
cAMP
cM
COMT
COMTD1
DNA
ENCODE
Gb
HPLC
IBD
indel
kg
kb
LD
LOD
Mb
mtDNA
RNA
SNP
TSHR
TYR
TYRP1
TYRP2
QTL

agouti signaling protein

beta-carotene oxygenase 2
beta-carotene oxygenase 2
base pair
cyclic adenosine monophosphate
centi-Morgan
catechol-O-methyltransferase 1
catechol-O-methyltransferase domain containing 1
deoxyribonucleic acid
Encyclopedia of DNA Elements
gigabase (109 bases)
high-performance liquid chromatography
identical by descent
small insertion and deletion
kilogram
kilobase (1000 bases)
linkage disequilibrium
logarithm of odds
megabase (106 bases)
mitochondrial DNA
ribonucleic acid
single nucleotide polymorphism
thyroid stimulating hormone receptor
tyrosinase
tyrosinase related protein 1
tyrosinase related protein 2
quantitative trait locus

Introduction

Charles R. Darwin combined geological, geographical and biological observations together as support for his theory that all organisms on planet Earth
share a common ancestry and that organisms are constantly evolving through
natural selection (1). Throughout his published work, Darwin used the process of domestication of animals and plants as a proof-of-principle that natural selection exists and acts (1, 2). He argued that if man through selective
breeding (artificial selection) has altered the morphology, physiology and
behavior (phenotypes) of the domesticates compared to their wild ancestors,
then the selective forces of the environment must achieve similar changes in
wild organisms (natural selection), however at a much slower rate. Darwin
was so amazed by the phenotypic variation observed in domestic animals
that he even devoted some of his precious time writing a book about the
subject: The variation of plants and animals under domestication (1868).
Darwin could never explain how the variation in traits arose or how the
variation was inherited from parents to their offspring. The publication Versuche ber Pflanzenhybriden in 1866 by Gregor Mendel gave some answers to the latter question. However, it did take almost more than half a
century until Mendels laws of inheritance were combined with Darwins
theories of natural selection to form the modern theory of evolution, known
as Neo-Darwinism. A number of landmark discoveries, such as the structure
of DNA (3) and the unearthing and deciphering of the genetic code, were
required until we had a clear picture that the variation in phenotypic traits is
caused by mutations in the DNA sequence that makes up our heritable genetic material, the genome.
In the year 2000, the sequence of the human genome was presented by
two separate efforts; private and public consortiums directed by Craig Venter and Francis Collins respectively. The analyses of the genome sequences
were published the year after in the scientific journals Nature and Science (4,
5). This has been followed by a number of generated genome sequences for
different animal and plant species in the last ten years, with an enormous
increase in the recent past as a consequence of the launch of new DNA sequencing technologies. The current sequencing technologiesmassively
parallel sequencing or next-generation sequencing technologiescan read
~250 billion bases in a week, compared to 5 million bases in the same time
period in the year of 2000 (6). The big challenge for contemporary molecular
geneticists is thus not to generate DNA sequences, but to decipher the ge9

nome sequencei.e. to understand how the genetic material determines how

we appear and function which is also known as the genotype-phenotype
relationship.
Domestic animals are excellent model organisms for exploring genotypephenotype relationships due to the enormous phenotypic variation among
them. No other experimental organism displays so much variation in phenotypic appearance in comparison to domestic animals. Information about the
genetic mechanism underlying a phenotypic trait in domestic animals
leading to the identification of the causal genecan shed light on the function of the human orthologous genes. Humans do in fact share a lot of genes
with domestic animals, and homologous vertebrate genes do often exert the
same function.
In this thesis, I will present two papers where we have used the chicken as
a model to identify genes that control variation in carotenoid- and melaninbased pigmentation (Papers I and III). The two identified genes are excellent
candidates for explaining the variation in carotenoid- and melanin-based
pigmentation that is observed in many other vertebrate species. In paper II,
we utilize newly emerging technological breakthroughs in DNA sequencing
to identify genomic regions that may have been under selection during both
domestication and the subsequent specialization of the chicken. This helps to
decipher the genetic basis of chicken domestication and the following specialization of egg and meat producing chicken lines.

Mutations the Source of Genetic Variation

Mutation and selection are the driving forces of evolution. Mutation is the
random process that increases variation, while selection primarily decreases
the levels of variation in the population (7). Mutations are changes in the
DNA sequenceor RNA in the case of some virusesthat arise during
DNA replication, recombination of chromosomes in meiosis, and also
through genetic transposition. Mutations can be categorized on the basis of
how they affect the DNA as: single nucleotide polymorphisms, small insertions or deletions, structural variation and transpositions. Single nucleotide
polymorphism(s) (SNPs) are the most common type of genetic variation, and
are associated with errors made by polymerase during genome replication.
Due to their high genomic abundance and ease to score, SNPs have become
the primary genetic marker used for genetic mapping and comparative analyses. Small insertions and deletions (indels) arise also primarily during genome replication, whereas structural variants (large insertions and deletions,
duplications and inversions) are errors made in the recombination process
during meiosis. The last type of mutation, transposition, is caused by jumping genes named transposable elements (TE) (8).

10

Functional Categorization of Mutations

The majority of mutations that arise do not have a functional consequence.
Those that can have a functional effect can be categorized as to whether they
affect the biochemical properties of the encoded protein or mature RNA:
coding mutations; or alter gene expression: cis-regulatory mutations (Figure
1) (9). Coding mutations fall within the exons of genes and alter the aminoacid sequence of the encoded protein or the nucleotide sequence of the mature RNA (10).
Cis-regulatory mutations occur in sequence elements of DNA that regulate the transcription of genes, namely cis-regulatory elements. The cisregulatory elementse.g. promoters, enhancers, silencers, and insulators
are scattered throughout genomes and via interaction with trans-acting factors (i.e. transcription factors) determine the timing and location (spatiotemporal) of gene transcription (11). The spatiotemporal transcription of a specific gene is under the control of a number of cis-regulatory elements, but
primarily those in close proximity to the gene, termed the cis-regulatory
module (CRM) (11). The regulation of gene expressionat the transcriptional levelvia the interaction between cis-regulatory elements and transacting factors, is fundamental to the differentiation of various cell-types and
therefore the development of complex organisms (11).

11

miRNA

Trans-factor

Target site

Cis-regulatory
elements

Promotor

Intron
Coding exon

Cis-regulatory Coding exon

element

Coding mutations

Cis-regulatory mutations

*
Deletion

Cis-regulatory
element

*
*
*

Figure 1. A figure depicting a few types of coding and cis-regulatory mutations. An

asterisk (*) represents the location of the mutation. Cis-regulatory elements can be
located within, upstream or downstream of the regulated gene. The mutations occurring in the genes encoding a miRNA or trans-factor may have secondary transacting effects on gene expression, but they are still considered to be coding mutations since they are located in coding sequences. Modified from Coyne and Hoekstra
(2007) (12).

Mutations in cis-regulatory elements have been suggested to play an important role in the evolution of phenotypic variation (13). King and Wilson
(1975) observed that human and chimpanzee showed surprisingly few protein sequence differences, and proposed that those alone could not account
for the phenotypic differences evident between the species. With the notion
of transcriptional regulation in minddiscovered by the bacterial geneticists
Jacob and Monod (1961) (14)they suggested that mutations that affect the
regulation of gene expression could account for the majority of the phenotypic differences observed between human and chimpanzee (15). A more
recent argument is that a cis-regulatory mutation is expected to be less pleiotropicaffect fewer phenotypescompared to a coding mutation, due to
the modular organization of the regulatory regions (13). A mutation in an
enhancer can, for instance, alter the expression of a gene in one tissue but
leave expression unaltered in other tissues. In contrast, a functional coding
mutation affects the function of the protein in all tissues in which it is expressed. This could lead to negative pleiotropic effects on many traits and
such mutations would thus be the object of purifying selection and perhaps
eliminated from the gene pool.

12

In paper I, we demonstrate that the yellow skin phenotype is caused by

one or several mutation(s) in cis-regulatory element(s) that alters the expression of the beta-carotene oxygenase 2 gene (BCDO2 or BCO2) in skin but
not in liver. A complete loss of function of BCO2 would most likely result in
negative pleiotropic effects on other traits, since the BCO2 protein has been
shown to play a protective role against carotenoid-induced mitochondrial
dysfunction (16).
Nevertheless, most of the mutations underlying phenotypic variation that
have been found to date have been coding (10). This can partly be explained
by the fact that once the gene controlling the trait has been determined, it is
much easier to identify a coding mutation than a regulatory mutation. The
genetic code makes it easy to predict if a mutation alters the function of the
protein by searching for non-synonymous, frameshift or nonsense mutations
(10). Mutations affecting transcription are much harder to identify due to our
limited knowledge regarding the location and function of cis-regulatory sequence elements. Functional and biochemical experiments are also often
required to confirm cis-regulatory mutations (9).

Identification of Sequence Elements Regulating Gene

Expression
Comparative genomicsthe comparison of genome sequences from different organisms with the aim to identify DNA sequences that are under evolutionary constraint and thus predicted to confer a biological functionhas
proven to be a successful approach for predicting the genomic location of
sequences that control gene expression (17). The genome sequence comparison between human and mouse revealed that more than 5% of the human
genome has been under purifying selection during the past 100 million years
and is thus probably functional (18). Surprisingly, only ~1.5% of the human
genome is protein coding, which implies that the majority of evolutionary
conserved sequences maintain other functions, such as gene regulation (18).
The study by Lindblad-Toh et al. (2011), which compared the genomes of 29
placental mammals, confirmed that ~5% of the human genome has undergone purifying selection. In addition they located constrained elements covering 4.2% of the human genomeof which more than two thirds were
found to reside within non-protein-coding regionsand experimental data
suggested that a large proportion of these are cis-regulatory elements (19).
Comparative analysis of genome sequences from species within the same
animal lineage can be used to identify sequences that are of functional importance for that specific animal lineage. For instance, the comparison of the
chicken, turkey and zebra finch genome sequences revealed that since they
shared a common ancestor, 8.6% of the chicken genome has been under

13

purifying selection (20). Some of these constrained sequences may be important in maintaining a number of traits that are specific for birds.
ENCODE, the Encyclopedia of DNA Elements, is a major research effort
with the aim to identify all functional elements in the human genome
through a combination of comparative genomic and experimental studies
(21). In 2007, only 1% of the human genome had been investigated, but the
effort continues and new data are released continuously (21). The aboveexemplified efforts in annotating genomes for functional elements will greatly simplify the detection of mutations that affect gene expression.

Domestication
Domestication refers to the human-directed process of making wild plants
and animals adapted to new environments and human needs. This adaption
has been achieved through selective breeding, or artificial selection, and has
resulted in an enrichment of mutations that control the phenotypes desired by
man (22).
The early domestication of animals was most likely unintentional, where
individuals carrying genetic variants predisposing them to be less afraid of
human approach were likely rewarded with food and protection (23). This
was later followed by intentional human selection for tame animals and other
desirable phenotypes, such as coat color variants (24, 25). Other typical phenotypic changes associated with animal domestication are as listed in the
review article by Jensen and Andersson (2005):
I
II
III
IV
V

External morphological changes (e.g. altered body size, fur and feather
color and growth pattern)
Internal morphological changes (e.g. reduced brain size)
Physiological changes (e.g. altered reproductive cycle and endocrine
response)
Developmental changes (e.g. earlier sexual maturity)
Behavioral changes (e.g. reduced fear, increased sociability and reduced
antipredator responses)

Many of these phenotypes frequently occur together as a block in numerous

different domesticated species, suggesting that they may represent general
adaptations to captivity and domestication (26). Belyaev and colleagues
showed in their classical selection experiment on silver foxes (Vulpes fulvus)
that many of the domestication phenotypesas described abovecould be
achieved by simply selecting against aggression (reviewed in Trut, 1999).
The reduction in aggression achieved after a few generations of selective
breeding was accompanied by a number of physiological and morphological
changes similar to the domestication phenotypes, such as coat color altera14

tions, dropping ears, shorter legs and tail, earlier sexual maturity, longer
mating seasons and alterations in skull morphology (27).
The rapid change in traits observed by Belyaev and colleagues, together
with the catalogue of shared domestication phenotypes among domestic
animals, have led some scientists to speculate that there are so-called domestication genes with large pleiotropic effects and that these created the
domestication phenotypes (28).
It has been proposed that thyroid hormones play an important role in animal domestication (29). Susan J. Crockfordthe main advocate of this theorycame to this conclusion after observing the similarities between domestication phenotypes and those associated with hypothyroidism (29). Crockford suggested that only a few changes of the regulation of thyroid hormones
during development, would be required to achieve many of the phenotypic
changes associated with the process of domestication. In paper II of this
thesis, we present data suggesting that the thyroid stimulating hormone receptor (TSHR)a key player in the regulation of thyroid hormone expressionhas played an important role in shaping the domestic chicken.
As described above, the domestication of animals has led to dramatic
changes in the phenotypic appearance of these individuals compared to their
wild ancestors (30). This long (~10 000 years) and intense selective breeding
has also created an enormous range of phenotypic variation within the domesticated species. The majority of these selected mutations have a favorable phenotypic effect on the selected trait (e.g. egg production or coat color)
but rarely a deleterious effect on other traits (22). Mutations with deleterious
effects are most often eliminated from the domestic animal breeding pool,
and so very few remain in the populations at large. The dog may be considered as an exception to that general rule, since selective breeding for specific
characteristics has resulted in pathological outcomeseither due to pleiotropism or linked deleterious mutationsthat have been accepted by the
breeders (31, 32).
The domestication of plants and animals is considered to be the most important development during the last 10 000 years of human history (33) and
was a prerequisite for the rise of civilization. Maintaining animals and plants
under our provision at the farm instead of hunting and gathering our food,
led to a population explosion due to the increased accessibility to food (33).
Permanent settlements could be established due to the decreased need for
migration to follow the seasonal shift in wild food supplies. It is no coincidence that inventions such as the wheel, metal tools, writing and advanced
social systems arose in the same areas as early domestication (33).

15

Domestication of the Chicken

The domestic chicken belongs to the genus of Gallus that comprises the four
wild species of jungle fowls: red jungle fowl (Gallus gallus), grey jungle
fowl (G. sonneratii), Ceylon jungle fowl (G. lafayetii) and green jungle fowl
(G. varius) (34). All four jungle fowl species inhabit different geographical
regions of South Asia and exhibit quite extensive differences in morphology
(Figure 2).

Figure 2. Panel A depicts a map of South Asia onto which the ranges of the four
species of jungle fowl are drawn. Panel B depicts a European domestic chicken.
Red, grey, blue and green regions represent the respective ranges of red, grey, Ceylon and green jungle fowls. Images of these birds are presented in panels C through
F respectively, within colored borders that correspond to the colors on the map. The
figure was adapted from Eriksson et al (2008).

The origin of the domestic chicken, in terms of geographical origin and

ancestry, has been an issue of debate since Darwin first proposed the Indian
red jungle fowl (Gallus gallus) as the single ancestor of all domestic chicken, and thus a monophyletic origin (2). He based his theory on the marked
morphological similarities between the domestic chicken and the red jungle
fowl, and the fact that crosses between the two generated fertile offspring,
while crosses between domestic chickens and the three other jungle fowl
species resulted in low survival of the chicks (2).
The alternative hypothesis for the ancestry of chickenthe multiplespecies or polyphyletic originwas first proposed by F.B. Hutt in his book
Genetics of the Fowl (1949). He claimed that potentially all four wild species
of jungle fowl may have contributed material to the genetic makeup of the
domestic chicken. He mainly based his hypothesis on the fact that a number
of phenotypes observed in domestic chickens are not found in the red jungle
fowl population, but are present in the other species of jungle fowl (35).

16

Early molecular genetic studies of mitochondrial DNA (mtDNA) supported the initial view that the red jungle fowl was the sole ancestor of the
chicken (36, 37). However, the study by Nishibori et al. (2005)which
compared sequences of mtDNA and five autosomal loci from domestic
chickens and all four jungle fowl speciessuggested that two additional
species of jungle fowl (grey and Ceylon jungle fowl) might hybridize with
the domestic chicken (38). Nevertheless, this study did not provide an answer to the question of whether they had contributed to the domestication of
chicken. In paper I of this thesis, we present the first conclusive genetic evidence supporting the theory of a hybrid or polyphyletic origin of the domestic chicken.
The earliest archeological findings of chicken remainsbones larger than
the red jungle fowlwere found in Southeast Asia and were estimated to be
almost 8000 years old (39). Additional archeological findingsbones and
art objects depicting chickens in the Indus Valley dated to about 2500
B.C.suggested that there have been multiple domestication events of the
chicken (34). The multiple geographical origin of the chicken is also supported by molecular genetic data that further implies that the majority of
European and Middle Eastern domestic chickens originate from India (40,
41). Further molecular studies present evidence that American, Oceanic and
African chickens also originated from India, suggesting that India was the
original platform for the worldwide dispersal of chicken (42-45).
Chicken was most likely initially domesticated for mainly cultural reasons
such as religion, art and entertainment and much later as a source of food
(34). This would consequently imply an early selection for increased body
size, an important feature in cockfighting, and for plumage variants, used as
decorative art. Most of the hundreds of chicken breeds currently in existence
were formed during the 1800s, when selection for fancy characterssuch as
plumage and skin color variation, amount of feathering, size of various external organsbecame high fashion among the royalties and the upper classes in Europe and America (34). Many Asian breeds were at that time imported to Europe, and very often hybridized with European breeds to create
new stocks with fascinating characters (34).
It was first during the 20th century when more systematic breeding
schemes focused on production traits were initiated (34). Chicken lines were
created with the sole task of producing either eggs or meatto overcome the
negative genetic and phenotypic relationship between reproduction and
growth (46). This has resulted in extreme differences in reproduction between the modern egg-layer chicken and its main wild ancestor, the red jungle fowl. The modern layer used in commercial egg production produces
more than 300 eggs per year, while a pure red jungle fowl lays around 5 eggs
once per year (47). The same extreme difference is observed in body weight
between the modern commercial broiler and the red jungle fowl, 4-5 kilogram (kg) versus 1 kg respectively (34). This great response to selective
17

breeding for egg and meat production has thus been achieved by sophisticated breeding programs, and is made possible due to the high genetic variability and large effective population sizes within the domestic chicken populations (46).
In paper II we, utilize massively parallel sequencing technology to search
for haplotypes in the chicken gene pool, which are thought to be responsible
for some of the phenotypic changes that have taken place during the domestication process, and the subsequent selection for specialized breeds used for
food production.
Chicken as Model Organism
The chicken has a long and proud history as a model organism. It was the
first animal in which the Mendelian inheritance of traits was demonstrated
(48), and scientific discoveries made with the chicken as an experimental
system in developmental biology (49), virology (50) and immunology (51)
have been awarded with Nobel Prizes.
The chicken possesses a number of features which make it particularly
suited to the study of genotype-phenotype relationships. Firstly, there is a
great diversity in phenotypessuch as plumage and skin color, body composition, production traitsproviding material for dissecting the genetic
mechanisms underlying phenotypic variation. Secondly, the chicken is easy
and cheap to breed, permitting the construction of large pedigree material
which segregates for the trait of interest (46). Third, the chicken genome has
a high recombination rate compared to many mammalian species (46, 52).
Fourth, many mutations underlying traits are shared between chicken breeds
due to gene flow and the recent history of breed creation (22). Fifth, there is
high genetic diversity within the chicken gene pool (53, 54). The last four
points provide excellent conditions for powerful genetic analyses to determine genotype-phenotype relationships.
Identifying the genes underlying some of the enormous variation in
chicken phenotypes is of great importance since it can be used to improve
the future breeding of domestic animals and also as it can provide information about the function of orthologous genes in vertebrates.

Pigmentation
One distinctive feature of all domestic animals is the huge variation in coloration among and within these individuals compared to their wild ancestors
(55). Selective breeding, since the early days of domestication (24, 25), has
achieved this phenomenon. The reason for that phenotypic selection might
have been to reduce camouflage in order to facilitate animal husbandry,
and/or to make the domesticated forms distinct from their wild ancestors, or
just as a human preference for novelty (24). The mutations controlling pig18

mentation variation were either present in the wild species prior to domestication or arose during domestication (55).
Pigmentation serves many different functions in animals, such as camouflage, mimicry, intraspecific communication, and protection against ultraviolet radiation (UVR) (56, 57). The mouse is the prime organism for the identification of genes controlling pigmentation due to the extensive mutagenesis
projects that have generated a tremendous number of pigment mutants. To
date (November, 2011), 171 genes have been identified which influence
pigmentation in the mouse in various ways, and more than 200 loci remain
to be mapped (58). Many of these genes have also been reported to affect
pigmentation in other vertebrate species, suggesting a high degree of conservation for the genes and molecular pathways controlling pigmentation in
vertebrates (55).
The chicken has also proven to be an excellent model organism for understanding the genetics of pigmentationillustrated by the number of pigmentation genes mapped in chicken (59-68). The identification of genes underlying plumage and skin color variants in the domestic chicken will provide
new insight about basic pigmentation biology.

Carotenoid-Based Pigmentation
Carotenoids are pigmentsyellow, orange or red in colorthat are synthesized by plants and microorganisms where they, for instance, facilitate photosynthesis and in plants protect chlorophyll from photo-damage (69). Animals lack the ability to synthesize carotenoids and must therefore acquire
them from their feed (69). Carotenoids serve several important functions in
vertebrate physiology. They act as precursors to vitamin A and retinoic acid
(70); anti-oxidative molecules (71); and photoprotectors of the human retina
(72).
Carotenoids are also important as pigments and add coloration to the tissue/organ where they are deposited, as observed in insects, fish, reptiles,
birds and mammals (69, 73). For instance, carotenoids are responsible for
the color of the egg yolk, lobster shell, salmon flesh and the flamingo plumage (69). In birds carotenoid-based pigmentation is known to act as an honest signal of the condition of a male carrier and functions as a sexually selected trait (74, 75). There is a trade-off between the allocation of carotenoids for colorful ornamentation versus the maintenance of carotenoids for
important physiological processese.g. as antioxidants in the immune system (74, 75)and only males in a good physical condition can afford to
allot carotenoids to ornamentation.
The genetic mechanisms involved in the metabolism and deposition of carotenoid pigments in animals were elusive until recently. In paper I, we present the first genetic mechanism underlying variation in carotenoid-based
pigmentation in animals. We conclude that a decreased expression of the
19

enzyme BCO2 (beta-carotene oxygenase 2) in the skinallowing the depositions of colorful carotenoidscauses the yellow skin and beak phenotype
in the domestic chicken.

Melanin-Based Pigmentation
The melanocyte is the cell type responsible for the synthesis of melanin
pigment in birds and mammals. Two types of melaninthe brown/black
eumelanin and the red/yellow pheomelaninare synthesized in the melanocyte-specific organelle, the melanosome (57). The melanin packed melanosome is later transferred to surrounding keratinocytes in the skin, eye, hair or
feather, where it provides color and photoprotection (57). It is the number,
size, cellular distribution and type of melanosomepheomelanosome or
eumelanonsomethat determines the color of the pigmented organ (76).
Research in the field of melanin-based pigmentation biology is extensive,
and so I will focus my review on the pathway that affects my work, namely
the synthesis of melaninalso know as melanogenesisor more specifically the production of pheomelanin pigment.
Melanogenesis
The production of melanin pigment is primarily initiated by the interaction
of external signaling proteinse.g. -melanocyte stimulating hormone (MSH) or agouti signaling protein (ASIP)to the melanocortin-1 receptor
(MC1R) that is located in the membrane of the melanocyte (77). The interaction between -MSH and MC1R leads to elevated levels of intracellular cyclic AMP (cAMP) that increases the transcription of the genes encoding
tyrosinase (Tyr) and the tyrosinase related protein 1 (Tyrp1) and 2
(Tyrp2/Dct) (78). Tyrosinase (Tyr) catalyzes the conversion of L-tyrosine or
L-dopa to dopaquinone, which is the precursor molecule to both eumelanin
and pheomelanin. The Tyrp1 and Tyrp2 enzymes catalyze reactions in the
biosynthesis of eumelanin (eumelanogenesis) with dopaquinone as the substrateleading to the production of the brown/black eumelanin pigment.
The interaction between ASIP and MC1R leads to lowered intracellular
levels of cAMP resulting in pheomelanin synthesis, which is also known as
pheomelanogenesis. The biosynthesis of pheomelanin requires the amino
acid cysteine in addition to dopaquinone (79). But unlike eumelanogenesis,
no additional enzymes seem to be required for pheomelanogenesis to proceed: pheomelanin is produced in the presence of dopaquinone and cysteine
(80). The potential lack of enzymes in pheomelanogenesis could explain
why only a scarce number of mice mutants (dwarf gray, lethal grey, grizzled,
subtle grey) (58) displaying a specific dilution or removal of pheomelanin
pigmentation have been identified. In contrast, numerous mice with mutations affecting only eumelanin pigmentation have been identified, and of
these 56 are caused by a mutated Tyrp1 gene (81).
20

Dopaquinonethe melanin precursor moleculeand intermediate molecules produced in melanogenesis are known to be toxic to the melanocyte
(57). To cope with those cytotoxic intermediates, melanogenesis takes place
in the closed environment of the melanosome (57). Occasionally some leakage to cytosol from the melanosome occurs, and protective mechanisms have
evolved to incapacitate the cytotoxic intermediates (57). One example of
such a mechanism is the O-methylation of the melanogenic intermediates
mediated by catechol-O-methyltransferase (COMT), leading to a reduced
cytotoxicity of the intermediates (57).
In paper III of this thesis, we propose that catechol-O-methyltransferase
domain containing 1 protein (COMTD1) is affecting pheomelanin pigmentation either; by the detoxification of cytotoxic pheomelanin intermediates
leaked from the melanosome; or by altering the cellular levels of cysteine; or
by acting as a regulator of pheomelanogenesisthis is because a frameshift
mutation in COMTD1 is underlying the Inhibitor of Gold phenotype in
chicken.

The Chicken Genome

The draft genome sequence of a single red jungle fowl female (UCD 001)
and the associated analysis was published in 2004 (52). The rationale for
sequencing a female bird was in order to generate sequence for both the Z
and W sex chromosomes, since in birds the female is the heterogametic sex
(82). In addition to the sex chromosomes, the chicken karyotype is made up
of 38 autosomal chromosome pairs that are divided into 10 large macrochromosome pairs (chromosome 1-10) and 28 small microchromosome pairs
(chromosome 11-38) (82). The microchromosomes generally have a higher
G+C content, gene density and recombination rate compared to the macrochromosomes (52). High G+C content is normally associated with low sequence representation in genome assemblies, which is reflected by the lack
of sequences assigned to ten of the 28 microchromosomes in the current
genome assembly (version 2.1, Washington University).
The recombination rate varies between 2.5 and 21 cM Mb-1 across the
chicken genome, but is on average higher compared to human (~1 cM Mb-1)
and mouse (~0.5 cM Mb-1) (52). The number of genes in the chicken genome
is estimated at 20 00023 000 of which 60% have a single human orthologue
(52). The haploid size of the chicken genome is around 1x109 basepairs (bp),
which is almost one third of most mammalian genomes (83). This size difference is likely due to lower numbers of interspersed repeats, segmental
duplications and pseudogenes (52). Small genome size is a feature observed
in most bird species and has been proposed to be an adaptation to the higher
rate of oxidative metabolism required to meet the energetic demand of flight
(83).
21

In order to assess the nucleotide diversity between and within domestic

chicken breeds, low coverage genome sequences (0.25 X) from three domestic chicken breeds (a Chinese silkie chicken, a broiler and a layer), were
generated by random-shotgun sequencing (53). The study identified 2.8 million single nucleotide polymorphisms (SNPs) and the analyses revealed a
remarkably high nucleotide diversity between and within breeds; on average
5 SNPs were detected every 1000 base pair (bp) in the pairwise comparisons
between and within breeds. This was comparable to the same amount of
variation that was observed in the domestic breeds to red jungle fowl comparisons. In conjunction, this suggested that there had been a historically
large population size of domestic chicken and that a significant amount of
the genetic variation has been captured from the wild ancestors (53).

Methods for Genetic Mapping

Introduction
A major challenge in biology and medicine is the assignment of function to
the sequences in the genome, with the ultimate goal being to understand the
link between genotype and phenotype. This has proven to be very challenging in the case of quantitative traits, to which multiple loci with small effects, environmental effects and epistatic interactions between loci all contribute to the resultant phenotype. Monogenic traitsthose that are determined by a single locusare on the other hand relatively easy to determine
the genotype-phenotype relationship of, and there are numerous molecular
and bioinformatic tools available that can be utilized for such studies. In
paper I and III, we identify the genes responsible for two monogenic traits,
using some of the methods that are available for gene mapping.
The new generation of sequence technologies has opened up a number of
possibilities for determining genotype-phenotype relationships. For instance
it is now possible, and cost-effective, to use massively parallel sequencing
and pooled samples to estimate the allele frequencies at all polymorphic loci
in different populations (84). Bioinformatic analyses of these data sets can
reveal genomic regions showing signs of positive selectionas a reduced
degree of heterozygosityin phenotypically similar populations. These regions are likely enriched for adaptive genetic variants that control the phenotypic differences observed between the compared populations. The regions
can be used as candidate regions to be further scrutinized by follow-up experiments in order to determine the link between the selected locus and a
phenotype.
In paper II, we use bioinformaticson whole genome sequence data generated from different DNA pools of domestic chicken breeds and the red
jungle fowlto identify genomic regions that most likely have been under
22

positive selection during the domestication process and the subsequent specialization of meat or egg producing chicken lines.

Linkage Analysis
Linkage analysis is a method used to identify a chromosomal region harboring a gene that underlies a phenotypic trait of interest. In order to perform a
linkage analysis three prerequisites need to be fulfilled: (i) a pedigree material segregating for the trait, (ii) genotype data from polymorphic genetic
markers and (iii) phenotype data of the individuals in the pedigree material.
Linkage analysis utilizes the phenomena of genetic linkage to identify
genetic markers that co-segregate with the phenotype of interest in a pedigree material. Genetic linkage refers to the observation that the closer two
genetic loci physically are to each other, the greater chance that they will be
inherited togetheri.e. they will not be separated by recombination during
meiosis (85). The genetic distance between marker pairs is measured in centi-Morgans and can be directly inferred from the recombination fraction: 1%
recombination between markers equals 1 centi-Morgan. The recombination
fraction () between markers is calculated by dividing the number of recombinant gametes with the total number of informative meiosis. Markers
are considered linked if the recombination fraction between them is significantly less than 0.5.
When performing linkage analysis, the phenotype is treated as a marker
with a specific mode of inheritance, and the aim is to detect genetic markers,
of known genomic positions, that are inherited together with the phenotypic
trait. To each marker-phenotype combination, a logarithm of odds (LOD)
score is assigned to statistically test the chance of obtaining a similar outcome as if the markers were unlinked (=0.5) (85). A LOD score of 3 is
generally considered to be significant evidence of linkage and means that
there is a 1 in 1000 chance that the observed linkage would have occurred by
chance (85).

Identical-By-Descent Mapping
Identical-by-descent (IBD) mapping, or linkage-disequilibrium analysis, is a
natural follow-up to linkage mapping and is used to refine the position of a
locus controlling a phenotypic trait. IBD mapping has proven to be especially powerful for detecting genes underlying phenotypic traits in domestic
animals (22).
IBD mapping utilizes the fact that many mutations underlying selected
phenotypes within domestic animals have been inherited in a manner termed
identical-by-descent (IBD) from a common ancestor and are therefore shared
between populations of a domesticated species (30). The intense selection
for such mutations in domestic animals has created so-called selective
23

sweeps in populations selected for the same phenotypes. A selective sweep

is characterized by a loss of heterozygosity at the selected locus and nearby
linked polymorphic loci, due to genetic hitch-hiking (86). The size of region
affected by a selective sweep is determined by both the local recombination
rate and the number of generations it took for the favorable haplotype to
become fixed in the selected population (87).
IBD mapping aims to detect a minimum haplotype shared across individuals from different breeds selected for the same phenotype (Figure 3). This is
usually performed by genotyping a dense set of polymorphic markers in the
trait-associated regionacross a large set of animals that are segregating for
the phenotype of interestin order to identify a minimum haplotype that is
completely associated with the phenotype across breeds. It is also feasible to
use the massively parallel sequencing technique on pooled samples to detect
selected lociwhich we prove in paper II of this thesis.
IBD mapping provides a much higher resolution in trait mapping than
linkage analysis, because the size of the associated region is based on many
historical recombination events, whereas linkage analysis is based on recombination events occurring within the pedigree and is limited by the size
of the pedigree material.

24

g=0
Ig

ig

g=2
ig

g=4
ig

g=8

ig

g=n

Ig

ig

Ig

Figure 3. Identical-by-Descent Mapping. An allele (ig) with a phenotypic effect

arises by mutation (*) in generation 0 (g=0). The mutant allele (ig) will be in complete linkage disequilibrium (LD) with alleles at other loci on the same chromosome.
LD will decay as a consequence of recombination, as the mutant chromosome is
passed down through the generations, but will although persist between the mutant
allele and alleles at closely linked loci. In generation n, a sample of ig and Ig chromosomes are genotyped to determine the minimum haplotype that is shared among
all animals carrying the ig allele. The minimum shared haplotype (yellow box) is
assumed to harbor the casual mutation. The figure was kindly provided by Dr. A. R.
Hellstrm who modified it from Andersson and Georges (2004).

Mutation Detection
Mutation detection used to be a daunting task due to the time-consuming
processes of primer-design, PCR amplification and nucleotide sequence
determination of the trait-associated region. Today, using the latest sequencing technologies provided by companies such Illumina and Life Technolo25

gies, it is possible to generate sequence for the trait-associated region within

a week (88). The generated sequence reads are mapped against a reference
genome sequence and sequence variants (SNPs, indels and structural variants) can be identified using various bioinformatic tools. The detected polymorphisms may then be intersected with exons of protein coding genes and
investigated as to how they affect the encoded protein sequence using public
software e.g. Variant Effect Predictor (89) and SIFT (90). Polymorphisms
may also be intersected with sequences that have been identified to be under
evolutionary constraintsuggesting that they possess a function, e.g. cisregulatory elements. Candidate mutations can, in the case of domestic animals, be further scrutinized by experimentation if they are shared between
breeds that display the same phenotype but are absent in those breeds for
which the phenotype is also absent.

26

Aims of this thesis

The overall aim of this thesis was to utilize genetic and genomic methods to
develop our understanding of the genetic mechanisms underlying phenotypic
variation and consequently add function to vertebrate genes.
The specific aims were to:

Identify the casual genes for two pigmentation phenotypes, yellow skin
(W) and Inhibitor of Gold (IG), in the domestic chicken (Paper I and
III).

Utilize massively parallel sequencing technology to search for signatures

of positive selection during chicken domestication, and the subsequent
artificial selection for egg and meat producing chicken lines (Paper II).

27

Present Investigations

Mapping Monogenic Traits in Chicken

Background
The long period of intense phenotypic selection has created an enormous
resource of variation in phenotypes within domestic animals. This rich resource is ready to be utilized, with the modern techniques in gene mapping
and mutation detection, in order to explore the genetic mechanisms underlying the variation in phenotypic traits. Whereas major effortsboth financial
and timehave been put into understanding the genetic mechanisms underlying quantitative traits of agricultural significance in domestic animals (22),
very few resources have been allocated to study the genes behind the numerous monogenetic traits observed in domestic animals. This is somewhat surprising since considerable amounts of money are spent on mouse mutagenesis programs to develop mutants that are used to assign function to genes
(22). The great number of monogenetic traits in domestic animals is therefore an underutilized resource with a great potential to assign function to
vertebrate genes and to determine genotype-phenotype relationships (22).
Numerous genes underlying coat color variation in mouse have been
identified (58). This is primarily due to the ease of detecting coat color phenotypes in mutagenesis screens, in comparison to many other internal and
external phenotypes that require more effort to detect. In addition, pigmentation phenotypes are also easy to score within the pedigree materials that are
established to identify the genes underlying pigmentation phenotypes. The
chicken has also proven to be an excellent model organism in which to identify genes underlying variation in plumage and skin pigmentation (59-66,
68). In paper I and III of this thesis, we report the identification of two
genesnever previously associated with variation in pigmentationto control the pigmentation phenotypes yellow skin and Inhibitor of Gold in chicken.
Understanding the genetics behind pigmentation in model organisms is,
for instance, of great importance when it comes to understanding what goes
wrong in human pigmentation disorders and skin cancers.

29

Yellow Skin Phenotype

The yellow skin phenotype in chicken was one of the first traits in animals
described to exhibit Mendelian inheritance, and was at the same time
demonstrated to be caused by the recessive autosomal yellow skin (W*Y)
allele at the W locus (48). Birds homozygous for the yellow skin allele display both yellow skin of the legs and body, and also a yellow beak. This is
the result of depositions of yellow carotenoids that the birds acquire from
their feed (35). The dominant white skin (W*W) allele does not allow depositions of carotenoids in the skin and beak, but does not seem to affect carotenoid depositions in other tissues (35). Today yellow skin is a very common
phenotype, and nearly all chicken used in commercial meat and egg production in the Western world possess yellow skincorresponding to billions of
chickens.
The red jungle fowl (Gallus gallus) has long been considered to be the
sole wild ancestor of the domestic chicken (1, 2). This was however challenged by Hutt (1949), who proposed that as many as three additional jungle
fowl speciesthe grey (G. sonneratii), Ceylon (G. lafayeteii) and green (G.
varius) jungle fowlsmay have contributed to the genetic makeup of the
domestic chicken. He based his view on the observation that a number of
phenotypes in the chicken are either found in the grey, Ceylon, and green
jungle fowl, but missing in red jungle fowl (35)including yellow skin.
Early molecular genetic studies have supported the view of the red jungle
fowl as the sole ancestor (36, 37). However a study from 2005 proposed the
possibility of hybridization between chicken and grey and Ceylon jungle
fowl (38), but it did not answer the question as to if they contributed to
chicken domestication.
Inhibitor of Gold Phenotype
The Inhibitor of Gold (IG) or Cream phenotype is a relatively common phenotype amongst chicken and is characterized by the dilution of the red
pheomelanin pigmentation of the plumage. Taylor first described IG in 1932
to be caused by the autosomal recessive IG*IG allele (91), which was later
confirmed by Punnett in 1948 (92). The wild type alleleleaving the
pheomelanin pigmentation unaffectedwas later denoted IG*N.
The regulation of the biochemical pathway leading to the production of
the red/yellow pheomelanin pigmentation is considered to be less understood
than the production of the black/brown eumelanin pigmentation. Only a few
mutations in the genes Slc7a11, Ggt, Ostm1 and SLC45A2 have been associated with a specific dilution of pheomelanin pigmentation (60, 93-95). Identification of the gene responsible for the Inhibitor of Gold phenotype in
chicken would thus provide a primer to decipher the biochemical regulation
of the synthesis of pheomelanin pigment. This is of great importance since
30

pheomelanin is thought to be involved in the development of UV-induced

skin cancer due to its phototoxic properties (96)however this is under debate (97).

Results and Discussion

Paper I Identification of the yellow skin gene reveals a hybrid origin of
the domestic chicken
In paper I, we demonstrate that the yellow skin phenotype is caused by one
or several tissue-specific cis-regulatory mutation(s) that leads to a reduced
expression of BCO2 (formerly BCDO2) in the skin, but not in liver. Surprisingly, we also demonstrate that the yellow skin allele most likely originates
from the grey jungle fowl and not from red jungle fowl, and thus present the
first evidence of a hybrid origin of the domestic chicken.
A combination of linkage analysis and IBD mapping across breeds with
the yellow skin phenotype was used to identify the gene responsible for yellow skin. Linkage analysis was performed using a backcross pedigreematerial (W/YY/Y), comprised of 91 individuals, that was provided by Hendrix Genetic B.V. The yellow skin locuspreviously assigned to chicken
chromosome 24 (98)was found to be in linkage with a SNP on chromosome 24 at nucleotide position 5,237,523 (LOD score=16.4; recombination
fraction=6.9%). The BCO2 gene located in proximity to the linked SNP
stood out as a candidate for yellow skin, as BCO2 encodes the enzyme betacarotene oxygenase 2 that had been reported to turn colorful carotenoids into
colorless apocarotenoids (99).
Partial resequencingfollowed by SNP genotypingof the BCO2 gene
in birds representing breeds segregating for yellow skin, revealed a 23.8
kilobase (kb) haplotype that all tested yellow skinned chicken were fixed for.
The haplotypeassumed to have been inherited identical-by-descent from a
common ancestorspanned the major part of BCO2 and the putative gene
BX935617.
Reverse Transcription PCR followed by pyrosequencing, of six heterozygous birds (Y/W), revealed that more than 90% of the BCO2 transcripts in
skin originated from the white skin (W*W) allele, whereas in liver, no difference in transcription from the Y and W alleles was detected. This suggests
the presence of one or several cis-regulatory mutation(s) affecting the transcription of BCO2 in skin, but not in liver. Consequently, we postulated that
both genotypes take up colorful carotenoids in the skin, but in white skinned
birds the carotenoids are degraded to colorless apocarotenoids by the action
of the BCO2 enzyme. After this study was published, mutations in BCO2
have been reported to be associated with increased carotenoid levels in: cow
milk (100, 101); cow fat (101); and sheep fat (73).

31

Sequence comparisons between the yellow skin and wild type alleles revealed the surprisingly high sequence difference of 0.81%, which was much
higher than the previously reported genome average of 0.5% (53). This observation, combined with the notion that both Ceylon and grey jungle fowl
have yellow/red legs, led us to sequence the 23.8 kb IBD region in the additional three jungle fowl species. The subsequent phylogenetic analysis revealed that the sequences clustered into two distinct clades (Figure 4). The
first comprised of BCO2 sequences from red jungle fowls and white skinned
(W/W) chickens, while the second clade comprised of sequences from yellow
skinned chickens and Ceylon and grey jungle fowls. We concluded that the
most plausible explanation to the observation was that the yellow skin allele
originates from the grey jungle fowlpossibly Ceylon jungle fowland
was introgressed into the chicken genome sometime during early domestication. Resequencing of the yellow skin interval in Chinese Shek-ki chickens
known to express the yellow skin phenotyperevealed that they carried
alternative yellow skin haplotype that was slightly different to yellow skin
haplotype found in Western breeds. The Chinese yellow skin haplotype did
however cluster with the Western yellow skin haplotype and BCO2 sequences from grey and Ceylon jungle fowls. The existence of two different yellow
skin haplotypes suggested that there have been multiple introgression events
of the yellow skin gene into the chicken gene pool.
Phylogenetic analyses of sequences of mtDNA and random nuclear loci
(38) derived from the four species of jungle fowl and yellow and white
skinned chickens, revealed a tight cluster of chicken and red jungle fowls
sequences separated from the other jungle fowl specieswhich suggested a
limited contribution of genetic material from the grey jungle fowl to the genetic makeup of domestic chicken.

32

Domes!c, Shek-ki
Grey jungle fowl
98

Ceylon jungle fowl

94

Yellow Skin

Grey jungle fowl

100

100

Domes!c, White Leghorn

Green jungle fowl
Red jungle fowl
100

Red jungle fowl

White Skin

63
54

Domes!c, Friesian
Red jungle fowl

Figure 4. A neighbor-joining tree depicting the relationships between sequences

derived from 23.8 kb of the BCO2 locus. Wild and domestic samples possessing
white and yellow skin clearly separate into two divergent clades. Node support values were generated from 1000 bootstrap replicates.

Since yellow skin is a very common phenotypeit has nearly replaced

white skin in commercial populationswe wanted to examine if the yellow
skin allele had any pleiotropic effects on other traits. Quantitative trait locus
(QTL) analyses on a pedigree material segregating at the yellow skin locus
revealed no significant association between the approximately 80 recorded
traitsincluding egg production and growthand yellow skin. We speculate
that the strong positive selection for yellow skin may be for cosmetic reasons
only, or that yellow skin had been used as an indicator of the egg-laying
capacity of a hen: egg-laying bleaches the skin color, as all acquired carotenoids are allocated to the egg-yolks (35), and this may have been utilized by
the early farmers to assess the length of a hens egg-laying period, a trait correlated with total egg-production.
Paper III A frameshift mutation in COMTD1 specifically dilutes
pheomelanin pigmentation in chicken
In paper III (manuscript), we present genetic data implying that the Inhibitor
of Gold phenotype is caused by a frameshift mutation in the previously uncharacterized
catechol-O-methyltransferase
domain
containing
1
(COMTD1) gene. This adds a new gene to the list of few genes that affect
the red pheomelanin pigmentation but which leave the brown/black
eumelanin pigmentation unaffected.
33

The Inhibitor of Gold locus was mappedby linkage analysis in a pedigree segregating for IGto a 3.58 megabase (Mb) region located between
the nucleotide positions 12,388,39915,970,174 on chicken chromosome 6.
No recombination was observed between genetic markers within the 3.58
Mb region and IG, which was a much larger non-recombinant region than
expected considering the size of the pedigree material (166 informative meiosis) and the high average recombination rate of the chicken genome (52).
However, a low recombination rate in the chromosomal region was also
observed in the consensus map for chicken (102), which made us conclude
that IG*IG allele is not associated with a large inversioninversions are
known to suppress recombination.
Using IBD mappingunder the assumption that most IG*IG chromosomes trace back to one common ancestorwe identified a shared haplotype
among IG birds from 5 different populations that spanned 438,921 bp. The
region harbored three ENSEMBL gene predictions corresponding to the
protein-coding genes COMTD1, C10ORF11 and ZNF503. A forth ENSEMBL gene prediction (ENSGALG00000024010) was also located in the
region but was only supported by a weak sequence similarity to a protein
encoded by the baboon herpes virus (Uniprot ID; Q9Q5K9). An even smaller haplotype associated with Inhibitor of Goldcontaining only
COMTD1was also determined. This IBD region was just supported by one
recombinant IG*IG chromosome and further analyses were performed on the
high confidence 439 kb minimum shared region.
To search for potential causal mutations for the IG phenotype, the whole
genome sequence of a bird homozygous IG*IG was generated, using nextgeneration Illumina sequencing technology. Polymorphic sites, including
single nucleotide polymorphisms, small insertions and deletions and large
structural variants, were identified and intersected with coding and evolutionary conserved sequences within the high confidence IBD regionin
which the causative mutation was supposed to be located. The only mutation
that fell in the coding or highly conserved sequences was a 2-bp-insertion in
the 5th exon of the COMTD1 gene. The insertion was predicted to alter the
reading frame of COMTD1 resulting in a premature stop codon.
Reverse Transcriptase PCR analysis of RNA samples derived from feather follicles representing the three genotypes at the IG locus, revealed the
presence of two different COMTD1 transcripts associated with the IG*IG
allele: one full-length transcript with the 2-bp insertion, and a second with
the 2-bp insertion that was lacking the 6th exon of COMTD1. The proteins
(Chicken_IG1 and Chicken_IG2) translated from the two IG*IG associated
transcripts, were both predicted to be completely or partially non-functional,
as both were lacking C-terminal sequences that were highly conserved
across vertebrate orthologs (Figure 5).

34

Figure 4
Chicken
Chicken_IG1
Chicken_IG2
Zebra_finch
G. Anole
Pufferfish
Dog
Human
Salmon
Mouse

. * :*:*.:*:**** * *
----MPLFSVPKEVAVGTAMLGVAFATGLLAGKRYPS-LLFGTSKPHKSIIGKSSPLWQYILDHSLREHPILKKL
----.................................................
.................................................. .....................
----.................................................. .....................
----........................M....L.....V..SPS---GM.R.NN..R.................
----....GIS..A.....V...VA-A.FF..R.SR.RFV...P.------.SPNL.Q..V......
....S.T..
----------------------------------------MSIL.S--HVGK--D..M..V.S..........G.
MSPPA.RL.V.AAL.L.S.A.GA...A...LGR.G.P---WRSRRERRLLPPEDS...R.L.SR.M......RS.
MTQPV.RL.V.PAL.L.S.A.GA......FLGR.C.P---WRGRREQCLLPPEDSR....L.SR.M......RS.
----MSFH.VSREALI.S.T.G.V.VA.YYIG..RST-FSMDIF.S--HSGA.DN..M..V.N......
.......
MAQPV.RL.I.AAL.L.S.A.GA.......LGK.W.P---W.SRRQERLLPPEDN.....L.SR.M......RS.
1.......10........20........30........40........50........60........70.....

Chicken
Chicken_IG1
Chicken_IG2
Zebra finch
G. Anole
Pufferfish
Dog
Human
Salmon
Mouse

:: * :
*:: :*:*::*** :**:***.:::*. ***.:* :**.:* * :::*::.
.::*:*.*
RLLTAEYPWGKMMVSCDQAQLMANLIKLIKAKKVIEVGVLTGYNALSMALALPDNGRVIACDINEDYAKIGKPLW
...........................................................................
...........................................................................
.....DH.S.R.....E........V..........I..F......N...V........................
KMV..DRRDKR....R.........AR..R.........F....T.N...VI..D.K.V...V..ES.NL...V.
..R.M.DSYNV...A.E.S.F....A..........I..Y....T........ED.ALV....S.E..N....F.
....L.Q.QGD..MT.E....L...AR......ALDL.TF...S..AL.....PA...VT.EV.PGPPEL.R...
....L.Q.QGDS.MT.E....L...AR..Q...ALDL.TF...S..AL.....AD...VT.EVDAQPPEL.R...
..R.MDDS.NFI.VAAE.S......AR......AI.I..Y....T..I..V...D.KMV....KDE.PN....F.
....L.Q.QGDS.MT E....L...AR......ALDL.TF...S..AL.....EA...VT.EVDAEPP.L.R.M.
...80........90.......100.......110.......120.......130.......140.......150

Chicken
Chicken_IG1
Chicken_IG2
Zebra finch
G. Anole
Pufferfish
Dog
Human
Salmon
Mouse

::*
:**:**::** :***:*:: *** ***...:**** .
***:.* *:: **::*: *: * *::*
KEAGVDHKIDLRIKPATQTLDELLAGGEAETFDFAFIDADKESYNEYYEKCLRLIKKGGIIAIDNVLRCGMVLKP
.................LKHLMNCWLVEKQKPLTLLSLMRIKKATMSTMKNVCAS
.................LKHL----- ---------------------------------------..........
.....E..........I......V.N..........................R.............FWN.K....
.....E..........S.......SN...G............G..D.........R...........LS.K..N.
.....EQ......Q..LK......S.........V......QN.DN....S.L.LR...VV......WG.D..N.
RQ.EEE...E..L...LE.......A...G...V.VV.....NCAA...R....LRS..VL.VLS..WR.E..Q.
RQ.EAE......L...LE.......A...G...V.VV.....NCSA...R....LRP...L.VLR..WR.K..Q.
.....AQ......Q..LK...D...N........V......AN.DN....S...VR........N..WS.K.VN.
.Q.E.EQ.....LQ..L........A...G...I.VV.....NCTA...R....LRP..VL.VLR.VWR.E..Q.
.......160.......170.......180.......190.......200.......210.......220.....

Chicken
Chicken_IG1
Chicken_IG2
Zebra finch
G. Anole
Pufferfish
Dog
Human
Salmon
Mouse

. . .: .**::: ** *: :*:: :.**::*.:*:

RKDDLATQSIHHLNEKLVRDARVNISMIPMGDGVTLVFKL
.............................M..........
......A....R.....L.........L.V......A...
ERYGGEV..M.E..K.IF..P..T....LM......A...
AP...D.VA.DK..K.LF..T..TL..LTV...L..AM..
QPR.VQARCVQN...RIL.....H..LL.L...L..A..I
P.G.V.AECVRN...RIR..V..Y..LL.L...L..A..I
AA..ID..A.DK..K..Y..V.I....LTV......A...
QPRNKTVECVRN...RIL.....Y.S.L.LD..LS.A..I
..230.......240.......250.......260.....

Figure 5. A protein sequence alignment of COMTD1 homologs in vertebrates,

including the wild type chicken COMTD1 and the two predicted protein sequences
encoded by the IG*IG allele. Periods (.) in the alignment indicate the same amino
acid as the full-length chicken master sequence and dashes () indicate gaps in the
alignment. An asterisk (*) indicates positions which have a single, fully conserved
residue. A colon (:) indicates conservation between amino acid groups of strongly
similar properties and a dot (.) indicates conservation between groups of weakly
similar properties according to the Gonnet Pam matrix made by Clustal X (103). The
arrows indicate residues that are in direct contact with the cofactor, S-AdenosylMethionine, based on the crystal structure model of COMTD1 (104). The black
square surrounding residues 1333 indicates the putative membrane-binding domain
while the black square spanning residues 62265 indicates the putative Omethyltransferase domain.

35

Homozygosity for the 2-bp insertion in COMTD1 was shown to be in

complete association with the Inhibitor of Gold phenotype in a pedigree
material segregating for IG. The 2-bp insertion was also demonstrated to be
in near complete linkage disequilibrium with IG*IG across breeds segregating for the phenotype. These results combined with the predicted severity of
the mutationa complete or partial loss-of-functionled us to propose that
the 2-bp insertion in COMTD1 is the causative mutation for the Inhibitor of
Gold phenotype.
We present three hypotheses that could explain how a non-functional
COMTD1 would result in specific dilution of pheomelanin pigmentation.
These theories are based on the known function of the well-characterized
catechol-O-methyltransferase (COMT) protein, which significantly resembles COMTD1 in terms of sequence and structure, including the predicted
O-methyltransferase domain.
We first propose that the COMTD1 protein may have the ability to alter
cysteine homeostasiscysteine is required for the synthesis of
pheomelanindue to its predicted function as an O-methyltransferase. Omethyltransferases are known to catalyze the transfer of a methyl group from
S-Adenosyl-Methionine (SAM) to a variety of substrates (105). This transfer
results in the conversion of S-Adenosyl-Methionine to S-AdenosylHomocysteine (SAH), which is an intermediate molecule in the cysteine
biosynthesis pathway (106). A non-functional COMTD1 protein would thus
potentially decrease the levels of cysteinedue to reduced SAM to SAH
conversionand that could subsequently affect the production of
pheomelanin.
The second hypothesis is that the COMTD1 protein regulates one of the
many reactions in the pheomelanin biosynthesis pathway. The COMT enzyme has been suggested to act as a regulator of melanogenesis due to its
ability to O-methylate L-dopa, which is a precursor molecule to melanin
(107, 108). We therefore speculate that the COMTD1 protein may act as a
pheomelanin synthesis specific pathway regulator.
The last theory we put forward is that COMTD1 exerts a protective role
against the cytotoxic metabolites produced during pheomelanongenesis. The
actual production of melaninboth eu- and pheomelaninhas been shown
to be cytotoxic due to the production of toxic intermediates that occasionally
leak from the melanosome into the cytosol (57). COMT has been proposed
to be involved in the detoxification process of these cytotoxic melanogeneic
intermediates (107) and COMTD1 may act in a similar manner during
pheomelanogenesis, but lack specificity to the intermediates produced as a
result of eumelanogenesis.

36

Signatures of Selection in the Chicken Genome

Background
A selective sweepcharacterized by a reduced degree of nucleotide variation at linked loci in a gene pool of interestis the product of strong positive
selection. Selective sweeps arise when a favorable allele at a genetic locus,
and hitchhiking neutral variants at neighboring loci, reach fixation or close
to fixation in a population (86). This can occur as a consequence of strong
natural or artificial selection and the size of the selective sweep region depends on the local recombination rate and the number of generations it took
for the selected allele to reach fixation.
Selective sweeps are a common feature in populations of domestic animals due to both the strong and recent selection for mutations with phenotypic effectsexemplified by the selective sweeps at the IGF2 locus in lean
pigs (109) and at the BCO2 locus in yellow skinned chickens (64). In 2004,
it was proposed that low-coverage whole-genome-sequences of domestic
animalsselected for the same purposecould be used to identify selective
sweeps on a genome-wide basis (30).
In paper II of this thesis, we describe the use of next-generation sequencing technology on pools of DNA samplesrepresenting different domestic
chicken populations and red jungle fowl populationsto identify haplotypes
that may have undergone selective sweeps during chicken domestication and
the latter specialization of egg and meat producing chicken lines. These haplotypes may contain alleles that could explain: how the domestic chicken can
produce up to 300 eggs per year in comparison to its major ancestor, the red
jungle fowl that lay a single clutch of 5 eggs per year (47); or how the commercial broiler easily can weigh 45 kg compared to the ~1 kg of the red
jungle fowl (34).
The chicken was selected for this experiment due to its small genome size
(~1Gb) and high neutral variation rate (~5SNPs/kb) within and between
chicken populations. For this study four broiler chicken lines were selected:
two commercial broiler lines collected in the AvianDiv project (110) and the
High and Low growth line (111) that have been divergently selected for
body weight at 8 weeks for more than 50 years. In addition, four egg-layer
chicken lines were selected: two White Leghorn populations, the OS-line
(112) and a commercial Rhode Island Red line (110). We also collected
samples from two red jungle fowl zoo populations. For each population we
selected the DNA from 8-11 birds to pool prior to the resequencing on the
Applied Biosystems SOLiD platform.

37

Results and Discussion

Paper 2 Whole-genome resequencing reveals loci under selection
during chicken domestication.
In paper II, we identify a number of loci showing a low degree of heterozygosity in eight domestic chicken populations, and which likely are the products of positive selection during chicken domestication. These regions are
assumed to be enriched for functional variants that have facilitated the transformation of the wild red jungle fowl to the domestic chicken. We also detected regions that have been under positive selection during the formation
of egg and meat producing chicken lines. In addition, we present a large
number of deletions in eight chicken populations, some of which may have
functional importance. The putative selective sweep regions provide an important raw material that can be used to investigate the differences observed
between the red jungle fowl and domestic chickens, and may also shed light
onto the genetic basis of the domestication process.
The whole-genome sequences were generated from the pools of DNA
samples representing eight different domestic chicken populations and a pool
of red jungle fowls. The genome sequence of the red jungle fowl (UCD 001)
birdpreviously used to generate the chicken genome assembly (52)was
also resequenced to function as a quality control for the sequence data generated with the SOLiD technology (Applied Biosystems, U.S.A). The short
35-bp sequences were aligned to the reference genome sequence and single
nucleotide polymorphisms (SNPs) were identified. Rigorous quality
checksincluding the experimental verification of 318 SNPs using an Illumina GoldenGate assaysuggested that the majority of the identified ~7,5
million SNP loci were true variants.
Next, the allele counts at all polymorphic sites were determined in three
different groups (breed pools) selected to represent the same purpose: all
eight domestic lines, three egg-layer lines (WL-A, WL-B and RIR) and two
broiler lines (CB-1 and CB-2). The rationale behind this move was to minimize the risk of detecting genomic regions fixed due to genetic drift. Putative selective sweeps were identified by calculating the pooled heterozygosity (Hp) in 40 kb windowsusing the allele counts within the three breed
poolssliding along the autosomes with the step size of 20 kb. The Hp values of each window were Z-transformed (ZHp) based on the observation that
the distributions of the Hp values in the three comparisons resembled normal
distributions. A window displaying a ZHp score of 6 or less was considered
to be a potential locus of selection, since these were in the extreme lower end
of the distributions.
The previously identified yellow skin allele at the BCO2 locus (64) was
used as a proof-of-principle for our approach to detect true selective
sweepssince all sequenced chicken populations were assumed to be fixed
for yellow skin. The 40-kb window that overlapped the yellow skin/BCO2
38

region showed the forth-lowest ZHp score (8.2) among all ZHp scores in the
all-domestic comparison, and thus confirmed the capacity of our approach to
detect true selective sweeps.
For the all-domestic comparison, a total of 58 windows representing 21
loci, displayed a ZHp score lower than 6. The putative selective sweep with
the lowest ZHp score (9.2) occurred at the locus encoding the thyroid
stimulating hormone receptor (TSHR) on chromosome 5. This was a noteworthy discovery since TSHR has a well-established role in metabolic regulation and reproduction (113-116), changes in which are associated with the
domestication process (26). To examine the geographical distribution of the
selected TSHR haplotype, genotyping of eight SNP loci at the TSHR locus in
271 birds representing 36 different chicken populations of a worldwide distribution, was performed. This revealed an amazingly high homozygosity at
the TSHR locus: 264 of the 271 birds were homozygous for the TSHR sweep
haplotype and the remaining 7 were heterozygous (Figure 6). This extreme
fixation observed in distantly related chicken populations led us to propose
that TSHR may be a so-called domestication locus, and have played an important role in shaping domestic chicken.
To search for selected mutations, Sanger sequencing was utilized to cover
gaps in the TSHR sweep region. The non-synonymous mutation at nucleotide
position 43,250,347 on chromosome 5resulting in the glycine (Gly) to
arginine (Arg) substitution at residue 558 in the TSHR proteinwas considered a top candidate for the selective sweep at the TSHR locus. The glycine
at that position was found to be completely conserved among all known vertebrate TSHR sequences, which suggested that the substitution most likely
would alter the function of TSHR.

39

Figure 6. Allele frequencies of the selected TSHR haplotypepresented as pie

chartsin chicken populations grouped on their continental origin (Europe, Asia,
Australia, North America and Africa). The grey color of the pie charts represents the
selected domestic TSHR haplotype while black represents non-selected haplotypes.
The allele frequencies are taken from Rubin et al (2010) and are based on data from
271 birds representing 36 different populations.

A previous experiment aimed at identifying QTL for growth used an intercross between the red jungle fowl and White Leghorn (117). Even though
the TSHR locus was determined to segregate in the intercross, it did not
overlap with any of the QTL regions previously reported, suggesting that the
mutation is not affecting growth. Instead, we speculate that the domestic
TSHR allele potentially acts to reduce the regulation placed on seasonal reproduction. TSHR signaling has an established role in the regulation of photoperiodic control of reproduction (114-116), and the absence of strict reguwww.VectorOpenStock.com
lation of seasonal breeding is in fact a common feature in domestic animals
(118).
Several of the putative sweep regions identified in the commercial broiler
lines occurred at genes with roles in obesity (TBC1D1), growth (IGF1,
INSR), and regulation of appetite (PMCH). This was anticipated since broiler
chickens have primarily been selected for muscle growth. An additional 5
putative selected loci with ZHp scores less than 6 were also identified in the
broiler lines.
We also searched for loss-of-function mutationsnonsense mutations
and deletionsin the sequence data, since these have been proposed as an
important factor in rapid evolution (119). Very little evidence was found that
the selection of loss-of-function mutations had been important during the
domestication process of chicken. Out of 1300 identified deletions, only
seven affected coding sequences, and two of those were considered to be
functional. The first deleted sequence in the growth hormone receptor (GHR)
gene, which was previously described to cause sex-linked dwarfism (120).
40

The second was a deletion in the uncharacterized gene SH3RF2 (SH3 domain containing ring finger 2). This deletion was considered to be especially
interesting since it sat within a QTL region for growth identified in the F2
generation of a cross between the High and Low growth lines. The High
growth line was fixed for the deletion, while it occurred at a low frequency
in the Low Growth line, and expression data from the hypothalamus showed
SH3RF2 expression in the Low growth line but not in the High growth line.
A QTL analysis with genotype data from the SH3RF2 locus in the F8 generation of the High and Low growth line intercross revealed a strong association between the deletion and increased growth. This suggests that the deletion in SH3RF2 is the causal mutation for the QTL on chromosome 20.

41

General Discussion and Future Perspectives

In papers I and III of this thesis, we present two genetic mechanisms underlying some of the variation in carotenoid and melanin pigmentation observed
in the domestic chicken. The genes identified to control the two monogenic
traits are excellent candidates for deciphering the genetic basis of the variation in carotenoid and pheomelanin pigmentation that is observed in many
other organisms. In paper II, we take one step closer to unraveling the genetic basis for the transformation of the wild red jungle fowl into the highly
productive domestic chicken. In combination, all three papers have advanced
our understanding of the genetic mechanisms underlying phenotypic variation.
The yellow skin study (Paper I) provides the first evidence of a hybrid
origin of the domestic chickenas proposed by Hutt in 1949. Despite that
the amount of genetic introgression from the grey jungle fowl seems to be
limited, it would be worthwhile to explore the extent of introgression from
the grey jungle fowl and the additional two jungle fowl species on a genomewide basis. This may be achieved by generating the genome sequences for
the four jungle fowl species, and a set of chicken breeds, and searching for
genomic regions where the chicken sequences cluster tightly with one of the
jungle fowls, rather than red jungle fowl. Such regions are thought to be
enriched for functional alleles since they have not been removed from the
chicken population through the extensive backcrossing that most likely took
place after the introgression. These alleles would probably also exert large
phenotypic effects on the selected traits, if we assume that the introgression
occurred early during domestication and before the worldwide dispersal of
chicken, when visual selection was the only selection method present. Mutations with small phenotypic effects on traits such as growth and reproduction
have probably first increased in frequency within the last 200 years or so,
when more advanced methods for phenotypic selection were developed.
This study also provides insight into the molecular mechanisms underlying variation in carotenoid pigmentation between two distinct wild species
namely the red and grey jungle fowlsand thus sheds some light on the
evolution of carotenoid-based pigmentation in wild species. Our study did in
fact compare sequences from two different species even though the sequences were segregating in the gene pool of a single species. With the fact that
grey jungle fowl has yellow/red skin and red jungle fowl white skin, we can
be quite confident that the mutation controlling the variation in carotenoid43

pigmentation arose prior to domestication. However, we are not able to determine if it is yellow or white skin that represents the ancestral statethere
is no reliable information of leg color of the closely related green jungle
fowl. It is quite remarkable that a study aimed at to identify the gene responsible for a phenotype observed within a single species, led to the detection of
the genetic mechanism underlying variation in a phenotypic trait, observed
between two distinct species.
The BCO2 gene is hence an excellent candidate for examining the genetic
mechanisms underlying variation in carotenoid-based pigmentation in other
species. We have recently started a projectin collaboration with evolutionary biologiststo test if BCO2 underlies the variation in beak color (yellow
or white/pink) observed in another bird species.
Our study also adds support to the view that cis-regulatory mutations have
played a significant role in the evolution of phenotypic variationthe yellow skin phenotype is caused by a cis-regulatory mutation affecting the expression of BCO2 in skin but not in liver. BCO2 has recently been shown to
be a mitochondrial protein with an important role in carotenoid homeostasis
and the protection of the mitochondria against carotenoid induced dysfunction (16). In addition Bco2 knockout mice display enlarged spleens and reduced testis size compared to wild-type mice after being fed a carotenoidrich diet (121). A complete loss-of-function mutation of BCO2 would most
likely confer deleterious effects on many traits and hence would be removed
from the gene pool. Interestingly, no adverse effects on health or development are associated with the complete loss-of-function (nonsense) mutations
of BCO2 causing increased carotenoid levels in cow milk (100, 101) and
yellow fat in cow (101) and sheep (73). A more thorough investigation of
numerous phenotypic traits for the different genotypes is required to truly
rule out the presence of deleterious effects associated with loss-of-function
mutations in BCO2 in these species.
Regarding the Inhibitor of Gold locus (Paper III), several functional experiments are required to develop our understanding of how the 2-bp insertion in COMTD1 results in reduced pheomelanin pigmentation. Characterization of the morphology and the number of melanocytes in feather follicles
derived from IG and wild type birds, would perhaps reveal if the COMTD1
protein exerts a protective role against the toxic intermediates produced during pheomelanogenesis. Whereas knock-down or over-expression experiments of COMTD1 in melanocyte cell lines, followed by the quantification
of pheomelanin using HPLC, would indicate if COMTD1 possesses a regulatory role in the synthesis of pheomelanin. To assess if COMTD1 is affecting
pheomelanin pigmentation, through regulating the levels of available cysteine for the pheomelanin synthesis, an appropriate experiment would be to
measure the cysteine levels in blood plasma from IG and wild type birds
(122).

44

It is interesting to speculate why COMTD1 has never previously been associated with pheomelanin pigmentation in other specieseven through
extensive mouse mutagenesis screens. One explanation could be that in other
species, a mutant COMTD1 would only have a subtle effect on pheomelanin
pigmentation and is thus not detected in the mouse mutagenesis programs.
Another explanation is that COMTD1 mutations are perhaps lethal due to
negative pleiotropic effects on essential physiological processes. The Inhibitor of Gold bird represents a unique resource in which to study the function
of COMTD1.
In paper II, we identified a number of loci in the chicken genome that
likely have been under selection during the domestication process of chicken. These regions are thus most likely enriched for alleles that can explain
some of the major phenotypic differences observed between the major wild
ancestor, red jungle fowl, and the domestic chicken. We also detected a
number of putative selected loci specific to meat or egg producing chicken
lines. The selected loci, found in broiler and layer chickens, may be utilized
to advance our knowledge of how energy metabolism, appetite and reproduction are regulated in vertebrate species.
First, further investigations in terms of genetic and functional experiments
are required to establish if sweep signals are the result of selection or genetic
drift. This may be followed up with experiments to determine the link between the selected haplotypes and phenotypes. One obvious effort that
would add support to the premise that selection has created the sweep signals, is if the selected mutations responsible for the sweep signatures were
found. For instance, non-synonymous single nucleotide polymorphisms
(nsSNP(s)) are fairly easy to detect and may be examined using SIFT (90) or
PolyPhen software (123), to predict the functional effect of the polymorphism. The nsSNPspredicted to have a functional effectcould be further
investigated using intercrosses between different chicken lines (e.g. egg layer/broiler or White Leghorn/red jungle fowl) if they were found to segregate
within these populations. QTL analyses with the SNP loci genotypes could
reveal if they are associated with any of the phenotypes collected from the
intercross. All putative sweep regionswith or without the presence of candidate mutationsshould be genotyped in intercrosses that are believed to
segregate at the sweep loci, in order to search for associations between the
sweep haplotypes and the recorded traits. In fact, such analyses are now being performed with broiler sweep loci in an intercross between egg layer and
broiler chickens.
The genes within, or in close proximity to, the putative sweep regions
should also be examined to see if they display differential expression between the appropriate populationsi.e. red jungle fowl vs. domestic and
layer vs. broilerthat may unearth if the selected mutations have a cisacting effect on gene expression. The combination of results from the abovedescribed analyses will perhaps provide the evidence required to decide
45

which of the putative selective sweeps that are the result of selection or genetic drift, and also possibly link the true selected loci to their phenotypes.
The supposed domestication locus at the TSHR gene should be followed
up with functional experiments and segregation analysis with data of phenotypic traits that are proposed to be affected by altered TSHR signaling. It is
tempting to speculate that the regulation of thyroid hormone signaling is
altered in other domesticated speciesresulting in many of phenotypes associated with the domestication process. This may be settled when the data
concerning the genetic basis of domestication in other animal species are
examined and made public.
Our study provides evidence that it is possible to detect loci that have
been under strong positive selection during chicken domestication using
next-generation sequencing technology on pools of DNA. The same should
now be possible for mammalian domesticated species with larger genomes
sizes than chicken, since the cost per sequenced base has dropped considerably in the last two years. It would also be interesting to utilize our approach
to detect signatures of selection in wild populations, in order to identify haplotypes controlling phenotypic traits that are important for adaptation to different environments.

46

Acknowledgements

This work was performed at the Department of Medical Biochemistry and

Microbiology, Uppsala, Sweden.
I would like to express my sincere gratitude to all the people that I have
had the pleasure to work with during my PhD studies. In particular I would
like to acknowledge the following people:
First, my supervisor Leif Andersson, for letting me be a part of so many
stimulating projects, and for guiding me during the current exciting period of
genome biology. It has been very inspiring to work with an excellent scientist such as yourself; you are a prime example of the correlation between
passion and success. A big thank you for the all the support you gave me as I
was struggling with finishing up a manuscript and my thesis at the same
time.
My co-supervisor Gran Hjlm, for sharing your knowledge of protein
biology and for keeping the local gym in excellent condition. Thanks for
ordering handsprit for me so that I could stay healthy during this busy
period.
Greger Larson, my second co-supervisor, for sharing your extensive
knowledge of domestication and phylogenetic analyseswhich resulted in
the yellow skin paperand for giving me access to your fine iTunes library.
Jennifer Meadows, colleague, for always having smart answers to my
questions, and for your extensive input on my thesis, despite being down
with a cold and fighting jet lag.
Carl-Johan Rubin, colleague, for your inspiring all-in attitude and precious help with the analyses of whole-genome sequence data.
All other excellent scientistsnone mentioned, none forgottenwho
have contributed to the findings presented in this thesis.
Anders Hellstrm, former colleague. We really had a good time in the lab
with our deep scientific discussions mixed with rubber band wars and Kejsarens kycklingsallad lunches. Im happy that you still are my friend, now
when you left the lab, and I wish you all the luck in south of Stockholm.
Lisa, Maria and Anders, my old office mates, for the hot dog sessions at
IKEA and the interesting and amusing discussions in the office.
All current and former members of the lab, for all your excellent support
and for making the years in the lab to an exciting time in my life. I will miss
our discussions on topics ranging from chicken breeding, HiFi equipment,

47

mushroom picking, rugby, Silvio Berlusconi, Chinese pop music, breastfeeding, beekeeping, to kitchen blenders.
Jag vill tacka min mentor Christer Harbeck som ftt mig att inse att
mycket spnnande vntar efter disputation. Samt fr att ha ptalat vikten av
planering och frberedelse infr stora prvningar, som att skriva avhandling
och att sedan frsvara den.
Jag vill ocks framfra ett stort tack till min omfattande familj i Upplands
Vsby, Leksand, Solna, Norrtlje, Tby, Julita, Tullinge, o.s.v. Jag vill speciellt tacka pappa, mamma, och mina brder fr all gldje och allt std ni
gett mig under rens gng. Ett gigantiskt tack till min mamma fr alla de
timmar du tragglade plutifikationstabellen och hglsning med decemberpojken Jonas. Tragglandet visade ju sig ge resultat!
Ett stort tack till alla mina vnner som utsttt min extra ptagliga frnvaro
det senaste halvret. Nu kommer det skert finnas mer tid fr ost- och vinkvllar och Halo sessioner.
Sist men absolut inte minst, vill jag tacka min underbara Anna, fr att du
ordnat med allt hemma, kpt mina julklappar, korrekturlst min avhandling
och till och med hunnit ordna en nobelmiddag t mig. Utan din matlagningskonst hade jag nog tappat ytterligare 5 kg i vikt, och utan dina lugnande ord
hade jag varit mycket mer stressad. Det ska bli underbart att nu f mer tid
med dig och jag ser framemot allt spnnande vi tillsammans har framfr oss
i livet.

48

References

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.

C. Darwin, On the origin of species by means of natural selection: or the

preservation of favoured races in the struggle for life. (John Murray, London,
ed. 1., 1859), pp. 502, 32 s.
C. Darwin, The variation of animals and plants under domestication. (O. Judd
& company, New York,, ed. Authorized, 1868), pp. 2 v.
J. D. Watson, F. H. Crick, Molecular structure of nucleic acids; a structure for
deoxyribose nucleic acid. Nature 171, 737 (Apr 25, 1953).
E. S. Lander et al., Initial sequencing and analysis of the human genome.
Nature 409, 860 (Feb 15, 2001).
J. C. Venter et al., The sequence of the human genome. Science 291, 1304
(Feb 16, 2001).
E. S. Lander, Initial impact of the sequencing of the human genome. Nature
470, 187 (Feb 10, 2011).
D. L. Hartl, A primer of population genetics. (Sinauer Associates, Sunderland,
Mass, ed. 3rd, 2000), pp. xvii, 221 p.
W. S. Klug, M. R. Cummings, Essentials of genetics. (Prentice Hall, Upper
Saddle River (N.J.), ed. 4th, 2002).
G. A. Wray et al., The evolution of transcriptional regulation in eukaryotes.
Mol Biol Evol 20, 1377 (Sep, 2003).
D. L. Stern, V. Orgogozo, The loci of evolution: how predictable is genetic
evolution? Evolution 62, 2155 (Sep, 2008).
E. H. Davidson, The regulatory genome : gene regulatory networks in
development and evolution. (Academic, Burlington, MA ; London, 2006), pp.
xi, 289 p.
J. A. Coyne, H. E. Hoekstra, Evolution of protein expression: new genes for a
new diet. Current biology : CB 17, R1014 (Dec 4, 2007).
G. A. Wray, The evolutionary significance of cis-regulatory mutations. Nature
reviews. Genetics 8, 206 (Mar, 2007).
F. Jacob, J. Monod, Genetic regulatory mechanisms in the synthesis of
proteins. J Mol Biol 3, 318 (Jun, 1961).
M. C. King, A. C. Wilson, Evolution at two levels in humans and
chimpanzees. Science 188, 107 (Apr 11, 1975).
J. Amengual et al., A mitochondrial enzyme degrades carotenoids and protects
against oxidative stress. The FASEB journal 25, 948 (Mar, 2011).
R. P. Alexander, G. Fang, J. Rozowsky, M. Snyder, M. B. Gerstein,
Annotating non-coding regions of the genome. Nature reviews. Genetics 11,
559 (Aug, 2010).
R. H. Waterston et al., Initial sequencing and comparative analysis of the
mouse genome. Nature 420, 520 (Dec 5, 2002).
K. Lindblad-Toh et al., A high-resolution map of human evolutionary
constraint using 29 mammals. Nature 478, 476 (Oct 27, 2011).

49

20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.

50

R. A. Dalloul et al., Multi-platform next-generation sequencing of the

domestic turkey (Meleagris gallopavo): genome assembly and analysis. PLoS
Biol 8, (2010).
E. Birney et al., Identification and analysis of functional elements in 1% of the
human genome by the ENCODE pilot project. Nature 447, 799 (Jun 14, 2007).
L. Andersson, in Comparative Genomics : Basic and Applied Research.
(Taylor & Francis Group, Boca Raton, 2007), pp. 22.
C. A. Driscoll, D. W. Macdonald, S. J. O'Brien, From wild animals to
domestic pets, an evolutionary view of domestication. Proc Natl Acad Sci U S
A 106 Suppl 1, 9971 (Jun 16, 2009).
M. Fang, G. Larson, H. S. Ribeiro, N. Li, L. Andersson, Contrasting mode of
evolution at a coat color locus in wild and domestic pigs. PLoS Genet 5,
e1000341 (Jan, 2009).
A. Ludwig et al., Coat color variation at the beginning of horse domestication.
Science 324, 485 (Apr 24, 2009).
P. Jensen, L. Andersson, Genomics meets ethology: a new route to
understanding domestication, behavior, and sustainability in animal breeding.
Ambio 34, 320 (Jun, 2005).
L. Trut, Early canid domestication: The farm-fox experiment. American
Scientist 87, (1999).
W. R. Stricklin, in Social behaviour in farm animals. (CABI Publishing,
Wallingford, 2001), pp. xvii, 406 p.
S. J. Crockford, in Human evolution through developmental change, N.
Minugh-Purvis, K. McNamara, Eds. (Johns Hopkins University Press,
Baltimore, 2002), pp. 122 153.
L. Andersson, M. Georges, Domestic-animal genomics: deciphering the
genetics of complex traits. Nature reviews 5, 202 (Mar, 2004).
N. H. Salmon Hillbertz et al., Duplication of FGF3, FGF4, FGF19 and
ORAOV1 causes hair ridge and predisposition to dermoid sinus in Ridgeback
dogs. Nat Genet 39, 1318 (Nov, 2007).
M. Olsson et al., A novel unstable duplication upstream of HAS2 predisposes
to a breed-defining skin phenotype and a periodic fever syndrome in Chinese
Shar-Pei dogs. PLoS Genet 7, e1001332 (Mar, 2011).
J. Diamond, Evolution, consequences and future of plant and animal
domestication. Nature 418, 700 (Aug 8, 2002).
R. D. Crawford, Poultry breeding and genetics. Developments in animal and
veterinary sciences, (Elsevier, Amsterdam ; New York, 1990), pp. 1123 s.
F. B. Hutt, Genetics of the fowl. (McGraw-Hill book co, New york, ed. 1st,
1949), pp. 590 s.
A. Fumihito et al., One subspecies of the red junglefowl (Gallus gallus gallus)
suffices as the matriarchic ancestor of all domestic breeds. Proc Natl Acad Sci
U S A 91, 12505 (Dec 20, 1994).
A. Fumihito et al., Monophyletic origin and unique dispersal patterns of
domestic fowls. Proc Natl Acad Sci U S A 93, 6792 (Jun 25, 1996).
M. Nishibori, T. Shimogiri, T. Hayashi, H. Yasue, Molecular evidence for
hybridization of species in the genus Gallus except for Gallus varius. Anim
Genet 36, 367 (Oct, 2005).
B. West, B.-X. Zhou, Did chickens go North? New evidence for
domestication. Journal of Archaeological Science 15, 515 (1988).
Y. P. Liu et al., Multiple maternal origins of chickens: out of the Asian
jungles. Mol Phylogenet Evol 38, 12 (Jan, 2006).

41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.

S. Kanginakudru, M. Metta, R. D. Jakati, J. Nagaraju, Genetic evidence from

Indian red jungle fowl corroborates multiple domestication of modern day
chicken. BMC Evol Biol 8, 174 (2008).
C. Berthouly-Salazar et al., Vietnamese chickens: a gate towards Asian genetic
diversity. BMC Genet 11, 53 (2010).
J. Gongora et al., Indo-European and Asian origins for Chilean and Pacific
chickens revealed by mtDNA. Proc Natl Acad Sci U S A 105, 10308 (Jul 29,
2008).
K. N. Dancause, M. G. Vilar, R. Steffy, J. K. Lum, Characterizing genetic
diversity of contemporary pacific chickens using mitochondrial DNA
analyses. PLoS ONE 6, e16843 (2011).
F. C. Muchadeyi et al., Mitochondrial DNA D-loop sequences suggest a
Southeast Asian and Indian origin of Zimbabwean village chickens. Anim
Genet 39, 615 (Dec, 2008).
P. B. Siegel, J. B. Dodgson, L. Andersson, Progress from chicken genetics to
the chicken genome. Poult Sci 85, 2050 (Dec, 2006).
P. A. Johnsgard, The pheasants of the world : biology and natural history.
(Smithsonian Institution Press, Washington, DC, ed. 2nd ed, 1999), pp. xvii,
398 p.
W. Bateson, Experiments with poultry. Repts. Evol. Comm. Roy. Soc. 1, 87
(1902).
S. Cohen, R. Levi-Montalcini, A Nerve Growth-Stimulating Factor Isolated
from Snake Venom. Proc Natl Acad Sci U S A 42, 571 (Sep, 1956).
P. Rous, A Sarcoma of the Fowl Transmissible by an Agent Separable from
the Tumor Cells. J Exp Med 13, 397 (Apr 1, 1911).
M. D. Cooper, D. A. Raymond, R. D. Peterson, M. A. South, R. A. Good, The
functions of the thymus system and the bursa system in the chicken. J Exp
Med 123, 75 (Jan 1, 1966).
International Chicken Genome Sequencing Consortium, Sequence and
comparative analysis of the chicken genome provide unique perspectives on
vertebrate evolution. Nature 432, 695 (Dec 9, 2004).
G. K. Wong et al., A genetic variation map for chicken with 2.8 million singlenucleotide polymorphisms. Nature 432, 717 (Dec 9, 2004).
C. J. Rubin et al., Whole-genome resequencing reveals loci under selection
during chicken domestication. Nature 464, 587 (Mar 25, 2010).
M. Cieslak, M. Reissmann, M. Hofreiter, A. Ludwig, Colours of
domestication. Biological reviews of the Cambridge Philosophical Society,
(Mar 28, 2011).
J. K. Hubbard, J. A. Uy, M. E. Hauber, H. E. Hoekstra, R. J. Safran,
Vertebrate pigmentation: from underlying genes to adaptive function. Trends
Genet 26, 231 (May, 2010).
J. J. Nordlund, The pigmentary system : physiology and pathophysiology.
(Blackwell, Malden, Mass. ; Oxford, ed. 2nd, 2006), pp. xviii, 1229 p., [84] p.
of plates.
L. Montoliu, Oetting, W. S. & Bennett, D. C. (European Society for Pigment
Cell Research, 2011).
S. Kerje, J. Lind, K. Schutz, P. Jensen, L. Andersson, Melanocortin 1-receptor
(MC1R) mutations are associated with plumage colour in chicken. Anim
Genet 34, 241 (Aug, 2003).
U. Gunnarsson et al., Mutations in SLC45A2 cause plumage color variation in
chicken and Japanese quail. Genetics 175, 867 (Feb, 2007).

51

61.
62.
63.
64.
65.
66.
67.
68.

69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.

52

U. Gunnarsson et al., The Dark brown plumage color in chickens is caused by

an 8.3-kb deletion upstream of SOX10. Pigment cell & melanoma research 24,
268 (Apr, 2011).
M. Vaez et al., A single point-mutation within the melanophilin gene causes
the lavender plumage colour dilution phenotype in the chicken. BMC Genet 9,
7 (2008).
A. R. Hellstrom et al., Sex-linked barring in chickens is controlled by the
CDKN2A /B tumour suppressor locus. Pigment Cell & Melanoma Research
23, 521 (Aug, 2010).
J. Eriksson et al., Identification of the yellow skin gene reveals a hybrid origin
of the domestic chicken. PLoS Genet 4, e1000010 (Feb, 2008).
S. Sato, T. Otake, C. Suzuki, J. Saburi, E. Kobayashi, Mapping of the
recessive white locus and analysis of the tyrosinase gene in chickens. Poult Sci
86, 2126 (Oct, 2007).
S. Kerje et al., The Dominant white, Dun and Smoky color variants in chicken
are associated with insertion/deletion polymorphisms in the PMEL17 gene.
Genetics 168, 1507 (Nov, 2004).
T. Tobita-Teramoto et al., Autosomal albino chicken mutation (ca/ca) deletes
hexanucleotide (-deltaGACTGG817) at a copper-binding site of the tyrosinase
gene. Poult Sci 79, 46 (Jan, 2000).
B. M. Dorshorst, A-M. Johansson, AM. Strmstedt, L. Pham, M-H. Chen, CF.
Hallbk, F. Ashwell, C. Andersson, L., A complex genomic rearrangement
involving the Endothelin 3 locus causes dermal hyperpigmentation in the
chicken. PLoS genetics in press, (2011, 2011).
G. A. Armstrong, J. E. Hearst, Carotenoids 2: Genetics and molecular biology
of carotenoid pigment biosynthesis. Faseb J 10, 228 (Feb, 1996).
J. M. Berg, J. L. Tymoczko, L. Stryer, Biochemistry. (W. H. Freeman, New
York, ed. 6th, 2007).
B. Demmig-Adams, W. W. Adams, 3rd, Antioxidants in photosynthesis and
human nutrition. Science 298, 2149 (Dec 13, 2002).
N. I. Krinsky, J. T. Landrum, R. A. Bone, Biologic mechanisms of the
protective role of lutein and zeaxanthin in the eye. Annu Rev Nutr 23, 171
(2003).
D. I. Vage, I. A. Boman, A nonsense mutation in the beta-carotene oxygenase
2 (BCO2) gene is tightly associated with accumulation of carotenoids in
adipose tissue in sheep (Ovis aries). BMC Genet 11, 10 (2010).
J. D. Blount, N. B. Metcalfe, T. R. Birkhead, P. F. Surai, Carotenoid
modulation of immune function and sexual attractiveness in zebra finches.
Science 300, 125 (Apr 4, 2003).
B. Faivre, A. Gregoire, M. Preault, F. Cezilly, G. Sorci, Immune activation
rapidly mirrored in a secondary sexual trait. Science 300, 103 (Apr 4, 2003).
D. Scherer, R. Kumar, Genetics of pigmentation in skin cancer--a review.
Mutat Res 705, 141 (Oct, 2010).
C. Wasmeier, A. N. Hume, G. Bolasco, M. C. Seabra, Melanosomes at a
glance. J Cell Sci 121, 3995 (Dec 15, 2008).
W. P. Walker, T. M. Gunn, Shades of meaning: the pigment-type switching
system as a tool for discovery. Pigment Cell & Melanoma Research 23, 485
(Aug, 2010).
J. D. Simon, D. Peles, K. Wakamatsu, S. Ito, Current challenges in
understanding melanogenesis: bridging chemistry, biological control,
morphology, and function. Pigment Cell & Melanoma Research 22, 563 (Oct
1, 2009).

80.

S. Ito, K. Wakamatsu, Chemistry of mixed melanogenesis--pivotal roles of

dopaquinone. Photochem Photobiol 84, 582 (May-Jun, 2008).
81. J. A. Blake, C. J. Bult, J. A. Kadin, J. E. Richardson, J. T. Eppig, The Mouse
Genome Database (MGD): premier model organism resource for mammalian
genomics and genetics. Nucleic Acids Res 39, D842 (Jan, 2011).
82. L. Stevens, Genetics and evolution of the domestic fowl. (Cambridge
University Press, Cambridge, 2005), pp. xiii, 306 p.
83. A. L. Hughes, M. K. Hughes, Small genomes for better flyers. Nature 377, 391
(Oct 5, 1995).
84. L. Andersson, How selective sweeps in domestic animals provide new insight
into biological mechanisms. J Intern Med, (Sep 3, 2011).
85. T. Strachan, A. P. Read, D. J. Matthes, Human molecular genetics 2. (BIOS
Scientific, Oxford, ed. 2nd, 1999), pp. xxiii, 576 p.
86. J. M. Smith, J. Haigh, The hitch-hiking effect of a favourable gene. Genet Res
23, 23 (Feb, 1974).
87. L. Andersson, Genetic dissection of phenotypic diversity in farm animals.
Nature reviews 2, 130 (Feb, 2001).
88. M. L. Metzker, Sequencing technologies - the next generation. Nature reviews.
Genetics 11, 31 (Jan, 2010).
89. W. McLaren et al., Deriving the consequences of genomic variants with the
Ensembl API and SNP Effect Predictor. Bioinformatics 26, 2069 (Aug 15,
2010).
90. P. C. Ng, S. Henikoff, Predicting deleterious amino acid substitutions.
Genome Res 11, 863 (May, 2001).
91. L. W. Taylor, in Sixth International Congress of Genetics (Ithaca, New York,
1932), pp. 197-199.
92. R. C. Punnett, Genetic studies in poultry; cream plumage. J Genet 48, 327
(Jan, 1948).
93. S. Chintala et al., Slc7a11 gene controls production of pheomelanin pigment
and proliferation of cultured cells. Proc Natl Acad Sci U S A 102, 10964 (Aug
2, 2005).
94. M. W. Lieberman et al., Growth retardation and cysteine deficiency in
gamma-glutamyl transpeptidase-deficient mice. Proc Natl Acad Sci U S A 93,
7923 (Jul 23, 1996).
95. N. Chalhoub et al., Grey-lethal mutation induces severe malignant autosomal
recessive osteopetrosis in mouse and human. Nat Med 9, 399 (Apr, 2003).
96. M. R. Chedekel, P. P. Agin, R. M. Sayre, Photochemistry of pheomelanin:
action spectrum for superoxide production. Photochem Photobiol 31, 553
(1980).
97. J. D. Simon, D. N. Peles, The red and the black. Acc Chem Res 43, 1452 (Nov
16, 2010).
98. M. Schmid et al., First report on chicken genes and chromosomes 2000.
Cytogenet Cell Genet 90, 169 (2000).
99. C. Kiefer et al., Identification and characterization of a mammalian enzyme
catalyzing the asymmetric oxidative cleavage of provitamin A. J Biol Chem
276, 14110 (Apr 27, 2001).
100. S. D. Berry et al., Mutation in bovine beta-carotene oxygenase 2 affects milk
color. Genetics 182, 923 (Jul, 2009).
101. R. Tian, W. S. Pitchford, C. A. Morris, N. G. Cullen, C. D. Bottema, Genetic
variation in the beta, beta-carotene-9', 10'-dioxygenase gene and association
with fat colour in bovine adipose tissue and milk. Anim Genet 41, 253 (Jun 1,
2010).

53

102. M. A. Groenen et al., A high-density SNP-based linkage map of the chicken

genome reveals sequence features correlated with recombination rate. Genome
Res 19, 510 (Mar, 2009).
103. J. D. Thompson, T. J. Gibson, F. Plewniak, F. Jeanmougin, D. G. Higgins, The
CLUSTAL_X windows interface: flexible strategies for multiple sequence
alignment aided by quality analysis tools. Nucleic Acids Res 25, 4876 (Dec
15, 1997).
104. J. W. Min, H. Zeng, H. Loppnau, P. Sundstrom, M. Arrowsmith, CH.
Edwards, AM. Bochkarev, A. Plotnikov, AN., The crystal structure of human
catechol-o-methyltransferase domain containing 1 in complex with sadenosyl-l-methionine., Protein Data Bank (2006).
105. T. T. Myhnen, P. T. Mnnist, Distribution and Functions of Catechol-OMethyltransferase Proteins: Do Recent Findings Change The Picture?
International Review of Neurobiology - Volume 95 95, 29 (Jan 1, 2010).
106. D. L. Nelson, A. L. Lehninger, M. M. Cox, Lehninger principles of
biochemistry. (Worth, New York, ed. 3rd, 2000), pp. xxix, 1231 p. in various
pagings.
107. N. Smit et al., O-methylation of L-dopa in melanin metabolism and the
presence of catechol-O-methyltransferase in melanocytes. Pigment Cell Res 7,
403 (Dec, 1994).
108. A. Slominski, D. J. Tobin, S. Shibahara, J. Wortsman, Melanin pigmentation
in mammalian skin and its hormonal regulation. Physiol Rev 84, 1155 (Oct,
2004).
109. A. S. Van Laere et al., A regulatory mutation in IGF2 causes a major QTL
effect on muscle growth in the pig. Nature 425, 832 (Oct 23, 2003).
110. J. Hillel et al., Biodiversity of 52 chicken populations assessed by
microsatellite typing of DNA pools. Genet Sel Evol 35, 533 (Sep-Oct, 2003).
111. E. A. Dunnington, P. B. Siegel, Long-term divergent selection for eight-week
body weight in white Plymouth rock chickens. Poult Sci 75, 1168 (Oct, 1996).
112. R. K. Cole, Hereditary hypothyroidism in the domestic fowl. Genetics 53,
1021 (Jun, 1966).
113. K. Dobney, G. Larson, Genetics and animal domestication : new windows on
an elusive process. Journal of zoology. 269, 261 (2006).
114. T. Yoshimura et al., Light-induced hormone conversion of T4 to T3 regulates
photoperiodic response of gonads in birds. Nature 426, 178 (Nov 13, 2003).
115. N. Nakao et al., Thyrotrophin in the pars tuberalis triggers photoperiodic
response. Nature 452, 317 (Mar 20, 2008).
116. E. A. Hanon et al., Ancestral TSH mechanism signals summer in a
photoperiodic mammal. Current biology : CB 18, 1147 (Aug 5, 2008).
117. S. Kerje et al., The twofold difference in adult size between the red junglefowl
and White Leghorn chickens is largely explained by a limited number of
QTLs. Anim Genet 34, 264 (Aug, 2003).
118. E. O. Price. Animal Domestication and Behavior (CABI Pub., Wallingford,
Oxon, UK ; New York, 2002), pp. 297 p.
119. M. V. Olson, When Less Is More: Gene Loss as an Engine of Evolutionary
Change. The American Journal of Human Genetics 64, 18 (1999).
120. S. K. Agarwal, L. A. Cogburn, J. Burnside, Dysfunctional growth hormone
receptor in a strain of sex-linked dwarf chicken: evidence for a mutation in the
intracellular domain. J Endocrinol 142, 427 (Sep, 1994).
121. N. A. Ford, S. K. Clinton, J. von Lintig, A. Wyss, J. W. Erdman, Jr., Loss of
carotene-9',10'-monooxygenase expression increases serum and tissue

54

lycopene concentrations in lycopene-fed mice. The Journal of nutrition 140,

2134 (Dec, 2010).
122. C. M. Pfeiffer, D. L. Huff, E. W. Gunter, Rapid and accurate HPLC assay for
plasma total homocysteine and cysteine in a clinical laboratory setting. Clin
Chem 45, 290 (Feb, 1999).
123. V. Ramensky, P. Bork, S. Sunyaev, Human non-synonymous SNPs: server
and survey. Nucleic Acids Res 30, 3894 (Sep 1, 2002).

55